James Angeles

I’m a Senior/Staff Machine Learning Engineer with 8+ years building and shipping production AI systems, focused on LLM platforms, multi-agent systems, and RAG. I take ML products end to end—from synthetic data and fine-tuning small models to running large models on GPUs and setting up MLOps and monitoring.

I’m strongest at turning big-frontier-model workflows into smaller, fine-tuned models that are faster and cheaper in production. I’ve delivered conversational search and applied-AI systems across healthcare documentation, knowledge-graph-driven retrieval, multi-turn reasoning, and agent orchestration.

At a healthcare startup client, I led generative AI system deployment for real-time medical documentation, helping reduce clinician note-taking time by over 63%. I also built conversational AI agent systems for a marketplace client, driving exploration and roadmap work for ranking, reasoning, and retrieval.

From multi-agent architectures (e.g., LangGraph-based workflows) to large-scale LLM inference (3B–600B parameters on H100 GPU clusters with vLLM/SGLang), I focus on measurable quality and reliability. I design full pipelines and observability—evaluation, A/B experimentation, tracing, and deployment automation—so teams can iterate quickly without breaking production.